1. Field of the Invention
This invention relates to a light scanning device and more particularly to a light scanning device wherein a plurality of light beams are irradiated, the plurality of irradiated light beams are deflected by a deflecting means and divided scanning is carried out on each scanning line by the deflected plural light beams, and a light scanning device for use in digital image forming apparatuses for electrophotography such as laser printers, laser copiers, and the like.
Further, this invention also relates to an optical device which is used in such image recording apparatus as laser printers, digital copiers and the like and in which light beam is swept according to image information so as to scan and expose a photoconductor, and a scanning method of the optical device, and more specifically to an optical device in which a single scanning line on the photoconductor is scanned in two divisions with two light beams at the same time and a scanning method of the optical device.
2. Description of the Related Art
In an ordinary optical scanning apparatus of the prior art, the face width of a polygon mirror is larger than the beam width in the main scanning direction of a light beam striking the polygon mirror. This face width is designed so as to cover the whole portion of the incident light beam regardless of any scanning angle (the so-called underfilled optical system).
In this underfilled optical system, as shown in FIG. 28, when an incident light beam having a beam width D0 in the main scanning direction is deflected by the polygon mirror, the beam width in the main scanning direction of a deflected light beam (deflection beam) is equal to the beam width of the incident light regardless of the scanning position. That is, assuming that the beam widths in the main scanning direction of the deflection beams leading to Start Of Scan (SOS), Center Of Scan (COS) and End Of Scan (EOS) are Ds, Dc, De respectively, D0=Ds=Dc=De is established. Consequently, the light volume and beam diameter (the effective diameter of a light spot converged on a photoconductor) at each scanning position are equalized thereby minimizing deterioration of image quality.
In recent years, the demand for higher recording speeds and resolutions has increased in such image recording apparatuses as laser beam printers, digital copiers, and the like using a light scanning device. To meet this demand for higher recording speeds and resolutions, a method may first be considered in which the time for a light beam to scan a single scanning line on the photoconductor is reduced by increasing the rotation speed of the polygon mirror.
However, there is a problem to be solved to achieve this idea. That is, usually the polygon mirror is rotated directly by a driving motor and currently, the upper limit of the rotation speed of the driving motor is 15,000 rpm (when ball bearings are used). However, this high speed motor is actually difficult to utilize because of the large increase in production costs. Even if pneumatic bearings are used, the limit is 40,000 rpm. Thus, there is an upper limit to the speeds and resolutions able to be obtained by increasing the rotation speed of the polygon mirror.
Increasing the number of deflecting faces of the polygon mirror can also be considered. However, if the number of deflecting faces increases, the diameter of the polygon mirror increases so that it is difficult to drive it with an ordinary driving motor. If, for example, under the underfilled optical system, it is intended to scan an A3 size sheet and ensure a beam diameter of about 60 .mu.m on the photoconductor, if the number of the faces of the polygon mirror is more than 10, the diameter of the polygon mirror exceeds 100 mm. To solve this problem, Japanese Patent Application Laid-Open (JP-A) No. 50-93719 has disclosed an overfilled optical system as a technology for avoiding an enlargement of the polygon mirror diameter (see FIG. 25).
As shown in FIG. 25, the light scanning device disclosed in the aforementioned patent comprises a light beam generating means 81, a modulation means 82A, a flat/convex cylindrical lens 86A having a curvature in a scanning direction, a focusing lens 88A, a polygon mirror 90A, an incline correcting cylindrical lens 92, and a photoconductor drum 94. According to the aforementioned patent, it is desirable that the number of the deflecting faces of the polygon mirror 90A is 20-30 and the scanning angle (.+-..alpha.) is .+-.12-18.degree..
According to the overfilled optical system, by expanding the beam width in the main scanning direction of light beam striking the polygon mirror beyond the face width of the polygon mirror as shown in FIG. 29, the diameter of the polygon mirror can be reduced, thereby making it possible to avoid an enlargement of the polygon mirror diameter even if the number of the deflecting faces thereof is increased.
However, if the number of the deflecting faces of the polygon mirror 90A is increased so as to increase the speed, the scanning angle at which the light beam is scanned by one deflecting face is inevitably decreased. Thus, the scanning width at a fixed distance from the polygon mirror 90A is decreased as the scanning angle is decreased. That is, to secure the same scanning width as in the prior art, the distance from the polygon mirror 90A to the photoconductor drum 94 needs to be increased, and the size of the light scanning device needs to be enlarged. For example, as is shown in FIG. 26, if it is intended to obtain a scanning width of 297 mm which is equivalent to A3 size paper, with the scanning angle (.+-..alpha.) being .+-.2.about.18.degree., the focal length f of the optical system exceeds 500 mm.
Although in the prior art, the central value of the beam diameter on the photoconductor drum 94 is assumed to be approximately 150 .mu.m, currently, as resolutions are intensified, this value has commonly come to approximately 60 .mu.m. Even in the overfilled optical system, if the number of the deflecting faces of the polygon mirror is 20-30 as shown in FIG. 27, the internal circle diameter of the polygon mirror exceeds 60 mm, and therefore this polygon mirror is difficult to rotate with a cheap motor. In FIG. 27, it is assumed that the scanning width is 297 mm and the beam diameter is 5 .mu.m.
As described above, even if the overfilled optical system is employed, there is a limit in the increase in speed, the increase in resolution, and the reduction in size able to be obtained.
Therefore, as an art in which a high speed and high resolution are realized while a small size is also attained, Japanese Patent Application Laid-Open (JP-A) No. 63-47718 has disclosed an optical device (light scanning device) in which divided scanning is performed on the surface of the photoconductor in the main scanning direction. This patent does not mention anything about overfilled optical systems.
In this optical device (light scanning device), the first half of an image area on the photoconductor is scanned with one laser beam and the second half thereof is scanned with an other laser beam. Because divided scanning is performed on the same scanning line with two laser beams, the number of the deflecting faces of the polygon mirror can be increased thereby increasing the print speed.
That is, in this optical device (light scanning device), as shown in FIG. 30, two light beams are projected from two laser beam sources onto the same point on the polygon mirror 9 so that the laser beams are perpendicular to the same deflecting face thereof and have different incident angles. Consequently, the divided regions 10, 11 on the photoconductor are scanned at the same time with two deflection beams from the polygon mirror 9. In this optical device, the difference in incident angle between two laser beam sources to a deflecting face is assumed to be .theta./2 while the entire scanning angle to a plane to be scanned of the photoconductor 3 is .theta.. Divided scanning is performed on the photoconductor at the scanning angle of .theta./2.
With this optical device, divided scanning is performed on the entire scanning plane at the same time with two light beams. Therefore, a higher speed and smaller size can be obtained as compared to an optical device which scans with a single light beam.
The diameter of the polygon mirror of this optical device depends on the width Dn of a light beam emitted from the polygon mirror which is set in such a manner that the beam formed on the photoconductor has the desired diameter. This emission beam width Dn is determined by the incident angle of the incident beam and incident beam width D0 for the underfilled optical system, and by the incident angle and face width (facet width) of the polygon mirror for the overfilled optical system.
The width Dn of emission beam of beam A 24A indicated by the dotted line leading to the scanning center position COS as shown in FIG. 31 is expressed as follows, where the incident angle of incident beam B 22B is .beta., the incident angle of beam A 21 is .beta.+2.alpha., the scanning angle is 2.alpha.(.+-..alpha.), the width of the incident beam to the polygon mirror is D0 and the face width in the main scanning direction of the polygon mirror deflecting face is FA;
for the underfilled optical system, EQU Dn=D0.times.COS((.beta.+2.alpha.)/2) (1)
for the overfilled optical system, EQU Dn=FA.times.COS ((.beta.+2.alpha.)/2) (2)
Meanwhile, referring to FIG. 31, the scanning range (-range) with the beam A is from the start of scan position (SOS) to the center of scan position (COS) and the scanning range with the beam B (t range) is from the center of scan position COS to the end of scan position EOS.
Because the aforementioned light scanning device performs divided scanning on the scanning line with two laser beams, a completed image is formed from a plurality of images created with respective laser beams.
Here, the respective laser beams are projected from a plurality of light sources located at different positions and arrive at the photoconductor through different paths. Thus, the mounting positions of the light sources, the positions of the parts composing the light sources, the positions of the optical systems mounted on the paths and the like are changed due to external factors such as temperature changes, vibration, impact and the like, so that the laser beam path may sometimes deviate from its predetermined path.
If the laser beam path deviates, discontinuities occur at the joints between the multiple images formed with respective laser beams, so that the image quality drops markedly.
Further, when an unevenness in the rotation speed of the polygon mirror occurs, the interval in the scanning line direction between dot positions of the multiple images created with the respective laser beams changes in each of the scanning lines. On the other hand, the first dot position is fixed since it is determined depending on a start timing signal from a predetermined sensor. Thus, the image quality at the joint between the multiple images drops markedly.
If 15.degree. and 45.degree. are substituted for .alpha. and .beta. in the aforementioned formulas(1), (2), it becomes necessary to set D0 or FA at about 1.26 times larger relative to Dn, so that there is a limit to the possible reductions in the diameter of the polygon mirror.
As shown in FIG. 31, for the center of scan position COS, it is desirable from the view point of image quality to form the beam diameters of the beam B 25 and beam A 24A projected in that direction and converged on the photoconductor with the same diameter. To achieve this, it is necessary to equalize the widths Dn of the beams A and B emitted from the polygon mirror projected at the center of scan position COS. For the underfilled optical system, the beam width Dn of the beam B 25 is expressed as follows because the incident angle is .beta.; EQU Dn=D0.times.COS (.beta./2) (3)
As evident from the formulas (1) and (3), to equalize the beam diameters of the beams A and B projected at the center of scan position COS in the underfilled optical system, the incident beam width D0 for beam A must be different to that for beam B.
For the overfilled optical system, the beam width Dn of the beam B 25 is expressed as follows, because the incident angle is .beta.; EQU Dn=FA.times.COS (.beta./2) (4)
As evident from the formulas (2), (4), to equalize the beam diameters of the beams A and B aimed at the center of scan position COS in the overfilled optical system, the face width of the polygon mirror must be different for beams A and B. However, this is physically impossible.
That is, according to the aforementioned conventional art, it is difficult or impossible to equalize the beam diameters of the beams A and B at the center of scan position COS which is the joint between the two beams. If the beam diameters differ at the center of scan position, an abrupt difference in line width occurs in images recorded at the center of scan position, thereby having a serious deleterious effect on the image quality.
Further, recently, higher speeds and resolutions have been demanded and even if the aforementioned disclosed art is used, the number of the deflecting faces of the polygon mirror needs to be more than 15. Thus, to avoid an enlargement of the polygon mirror diameter, there is no alternative but to employ the overfilled optical system.
If the overfilled optical system is applied in the aforementioned prior art, the following problems inherent in the overfilled optical system occur.
Because part of the light beam projected at the polygon mirror is cut off for use as a light beam in scanning, the FN0 (brightness in camera terminology) changes depending on the scanning angle (scanning position), so that the uniformity of beam diameter at the focusing position (in the vicinity of the photoconductor) worsens when linking therewith.
The FN0 mentioned here will be explained in detail. Assuming that the focal length of a focusing optical system for converging a laser beam deflected by the polygon mirror on the photoconductor is f, and the width (emission beam width) of a beam emitted from the polygon mirror is Dn, the FN0 is expressed as follows; EQU FN0=f/Dn (5)
Further, the beam diameter L.sub.b on the photoconductor is substantially proportional to the product of the beam wavelength .lambda. and FN0. Therefore, by using k for the proportional coefficient, the formula (5) can be expressed as follows; EQU L.sub.b =k.lambda..multidot.FN0=k.lambda..multidot.f/Dn (6)
As evident in the formula (6), the beam diameter L.sub.b is proportional to focal length f and the beam diameter L.sub.b is inversely proportional to the width Dn of the beam. That is, as the focal length f is elongated, if it is intended to be kept at the same beam diameter, Dn needs to be increased, so that the size of the polygon mirror is also increased.
As shown in FIG. 28, in the underfilled optical system, Dn is constant regardless of the scanning angle. On the other hand, in the overfilled optical system, as shown in FIG. 29, the emission beam width Dn changes depending on the scanning angle. That is, when the width of the beam leading to the start of scan position SOS is Ds, the width of the beam leading to the center of scan position COS is Dc and the width of beam leading to the end of scan position is De, the following relation exists; EQU Ds&lt;Dc&lt;De&lt;D0
According to the formulas (5), (6), the FN0 and beam diameter L.sub.b change depending on the scanning position on the photoconductor. This beam diameter L.sub.b changes at each scanning position by the factor 1/COS (*****) which is the inverse number of COS (*****) because a relation of Dn=FA.times.COS (*****) exists as indicated by the formula (4). Therefore, a ratio of the factor 1/COS (*****) or a ratio of FN0 is a parameter for indicating uniformity of the beam diameter L.sub.b.
As for the example shown in FIG. 31, the factor 1/COS (*****) is:
At the scanning position (SOS) with beam 23, 1/COS (.beta.+4.alpha.)/2 PA1 At the scanning position (COS) with beam 24A, 1/COS (.beta.+2.alpha.)/2 PA1 At the scanning position (COS) with beam 25, 1/COS (.beta./2) PA1 At the scanning position (EOS) with beam 26B, 1/COS (.beta.-2.alpha.)/2 PA1 at the scanning position (SOS) with beam 23, 1/COS (.beta.+4.alpha.)/2=1.643 PA1 at the scanning position (COS) with beam 24A, 1/COS (.beta.+2.alpha.)/2=1.260 PA1 at the scanning position (COS) with beam 25, 1/COS (.beta./2) =1.082 PA1 at the scanning position (EOS) with beam 26B, 1/COS (.beta.-2.alpha.)/2=1.009.
If the number of the deflecting faces of the polygon mirror is n=20, the angle of each face of the polygon mirror is 360.degree./20 =18.degree.. Thus, the rotation angle .+-..alpha./2 of the polygon mirror is less than .+-.9.degree. (.alpha. is less than 18.degree.).
Where .alpha. is 15.degree. and .beta. is 45.degree.;
Thus, the FN0 ratio between SOS and EOS is 1.643/1.009=1.628Therefore, if the beam diameter at the EOS is 60 .mu.m, the beam diameter at the SOS is about 100 .mu.m, which is an unacceptable level in terms of image quality. Further, at the center of scan position COS where two beams join together, there occurs a difference in terms of FN0 of 1.082 and 1.260, which, in, terms of beam diameter, are 60 .mu.m and 70 .mu.m. The difference of 10 .mu.m has a serious deleterious effect on the image quality resulting in an abrupt difference in line width at the center of scan position.
Further, because in the conventional overfilled optical system, the farther from the incident optical axis, the wider the removed part near the peak of Gausian distribution, the reduction of light volume at a side far from the incident optical axis becomes larger than the reduction of the beam diameter (see FIGS. 23, 24 described later). That is, together with the difference in beam diameter at the center of scan position, an abrupt difference in light volume occurs between the center of scan position and the sides far from the incident optical axis, having a serious deleterious effect on image quality.
Further, because the photoconductor is scanned with a plurality of laser beams when the aforementioned light scanning device is used, if light volumes at the joint between the adjacent scanning beams are not equal, stripes are formed at the joint between the adjacent scanning beams unlike scanning with a single laser beam, so that the formed image becomes fragmentary thereby deteriorating the image quality.
To solve this problem, in the light scanning device disclosed in Japanese Patent Application Laid-Open (JP-A) No. 58-127912, an area in which scanning beams optically overlap each other is provided for the joint area where the images are fragmented, and this area is assumed to be boundary area. In this boundary area, joints are set at random so as to make stripes which properly occur there apparently invisible.
Further, in the light scanning device disclosed in Japanese Patent Application Laid-Open (JP-A) No. 3-98066, an area in which scanning beams optically overlap each other is provided. Then, the exposure energy of one scanning beam is reduced and that of another scanning beam is increased, so that the total exposure amount in the overlapping area becomes an average value which is not largely different from other exposure ranges.
According to the above described inventions, as a method for correcting image distortions in the joint area between the exposure ranges, there have been proposed a method of changing the joint position and a method of averaging light volumes. When an image is formed on the photoconductor, if the image data which actually modulates the laser beam is taken into account, the method of changing the joint position givens rise to the problem that the relation between the exposure range to be scanned and the image data becomes complicated because the joints are randomly different. Meanwhile, if according to this method, changing of the joint position is carried out cyclically so as to make the relation between the exposure range to be scanned and image data less complicated, image distortion occurs in that same cycle.
In the method of averaging light volumes to correct image distortion in the joint area between the exposure ranges, averaging the light volumes is no problem, but the overlapping area is written twice depending on image data, so that the image blurs. This problem is particularly fatal in cases when a higher resolution is needed.
There is a common problem in both the method of changing the joint position and the method of averaging the light volumes. That is, when a semiconductor laser is used as the light source of the light scanning device, even if the optical outputs of multiple semiconductor lasers for use in scanning multiple exposure ranges have constantly equal light volumes, the following problem exists. Because the semiconductor laser has droop characteristics, the transient change in light volume which occurs at the joint between the divided exposure ranges cannot be corrected.
Namely, laser beams emitted from two semiconductor lasers scan separate exposure ranges across the joint there between. A laser beam for scanning the exposure range in front of the joint, of the two exposure ranges , of ten has constant light volume around the joint. However, a laser beam for sweeping the exposure range behind the joint has an excessive light volume which occurs at boot-up, resulting in the light volume of the laser beam automatically increasing because of the droop characteristics. Therefore, despite changing the joint positions at random and averaging the light volumes in the joint area, image distortion or stripe formation occurs.